RandomForest_new.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "b71e70b8",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import datetime\n",
    "import os\n",
    "import re\n",
    "import seaborn as sns\n",
    "from collections import Counter\n",
    "from datetime import datetime\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import train_test_split, StratifiedKFold, RandomizedSearchCV\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3d7843a6",
   "metadata": {},
   "source": [
    "RANDOM FOREST"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "5c3cfeb4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.00546968, -2.63575159, -0.08626786, ..., -0.49449405,\n",
       "        -0.73591438,  3.30735414],\n",
       "       [ 0.00546968, -2.63538219, -0.08626786, ..., -0.49449405,\n",
       "        -0.73591438,  3.30735414],\n",
       "       [ 0.00546968, -2.59997259, -0.08626786, ..., -0.49449405,\n",
       "        -0.73591438,  3.30735414],\n",
       "       ...,\n",
       "       [ 0.99466784,  1.87366874, -0.08626786, ...,  2.02226904,\n",
       "        -0.73591438, -0.30235649],\n",
       "       [ 0.99466784,  1.88971124, -0.08626786, ...,  2.02226904,\n",
       "        -0.73591438, -0.30235649],\n",
       "       [ 0.99466784,  1.90575374, -0.08626786, ...,  2.02226904,\n",
       "        -0.73591438, -0.30235649]])"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = np.load('X.npy')\n",
    "X"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "56729343",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0],\n",
       "       [0],\n",
       "       [0],\n",
       "       ...,\n",
       "       [0],\n",
       "       [0],\n",
       "       [0]])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y = np.load('y.npy')\n",
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "9f90e301",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1709857,)"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y = y.reshape(-1)\n",
    "y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ffac0945",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "7a3cb228",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0,\n",
       " 56,\n",
       " array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16,\n",
       "        17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,\n",
       "        34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50,\n",
       "        51, 52, 53, 54, 55, 56]))"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.min(y), np.max(y), np.unique(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "be38f27c",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "9fbfb59c",
   "metadata": {},
   "outputs": [],
   "source": [
    "skf = StratifiedKFold(n_splits=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "40589188",
   "metadata": {},
   "outputs": [],
   "source": [
    "rfc = RandomForestClassifier()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "77fbee56",
   "metadata": {},
   "outputs": [],
   "source": [
    "param_grid = { \n",
    "    'n_estimators': [100],\n",
    "    'max_features': ['auto'],\n",
    "    'max_depth' : [None],\n",
    "    'criterion' :['gini']\n",
    "}\n",
    "random_search = RandomizedSearchCV(rfc, param_grid, cv=skf, verbose=2, random_state=42, n_jobs = -1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "8b75f477",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 2 folds for each of 1 candidates, totalling 2 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.9/site-packages/sklearn/model_selection/_search.py:292: UserWarning: The total space of parameters 1 is smaller than n_iter=10. Running 1 iterations. For exhaustive searches, use GridSearchCV.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'mean_fit_time': array([85.52497005]), 'std_fit_time': array([0.69346094]), 'mean_score_time': array([29.34886956]), 'std_score_time': array([0.04427743]), 'param_n_estimators': masked_array(data=[100],\n",
      "             mask=[False],\n",
      "       fill_value='?',\n",
      "            dtype=object), 'param_max_features': masked_array(data=['auto'],\n",
      "             mask=[False],\n",
      "       fill_value='?',\n",
      "            dtype=object), 'param_max_depth': masked_array(data=[None],\n",
      "             mask=[False],\n",
      "       fill_value='?',\n",
      "            dtype=object), 'param_criterion': masked_array(data=['gini'],\n",
      "             mask=[False],\n",
      "       fill_value='?',\n",
      "            dtype=object), 'params': [{'n_estimators': 100, 'max_features': 'auto', 'max_depth': None, 'criterion': 'gini'}], 'split0_test_score': array([0.92673072]), 'split1_test_score': array([0.92549073]), 'mean_test_score': array([0.92611072]), 'std_test_score': array([0.00062]), 'rank_test_score': array([1], dtype=int32)}\n"
     ]
    }
   ],
   "source": [
    "random_search.fit(X_train, y_train)\n",
    "print(random_search.cv_results_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "dae5eea5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9261107239457991"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "random_search.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "2f0c6f63",
   "metadata": {},
   "outputs": [],
   "source": [
    "classifier = random_search.best_estimator_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "8505c9ae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9999724287512981"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classifier.score(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "bc51ba5b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.957269406072232"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classifier.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "56579cfd",
   "metadata": {},
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "'RandomizedSearchCV' object has no attribute 'scores_'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "Input \u001b[0;32mIn [17]\u001b[0m, in \u001b[0;36m<cell line: 1>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0m \u001b[43mrandom_search\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mscores_\u001b[49m\n",
      "\u001b[0;31mAttributeError\u001b[0m: 'RandomizedSearchCV' object has no attribute 'scores_'"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] END criterion=gini, max_depth=None, max_features=auto, n_estimators=100; total time= 1.9min\n",
      "[CV] END criterion=gini, max_depth=None, max_features=auto, n_estimators=100; total time= 1.9min\n"
     ]
    }
   ],
   "source": [
    "random_search.scores_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "b68b17ca",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 360x360 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#modelname.feature_importance_\n",
    "features = classifier.feature_importances_\n",
    "#plot\n",
    "fig, ax = plt.subplots()\n",
    "width = 0.4 # the width of the bars\n",
    "ind = np.arange(len(features)) # the x locations for the groups\n",
    "ax.barh(ind, features, width, color=\"green\")\n",
    "ax.set_yticks(ind+width/10)\n",
    "ax.set_yticklabels(range(len(features)), minor=False)\n",
    "plt.title(\"Feature importance in RandomForest Classifier\")\n",
    "plt.xlabel(\"Relative importance\")\n",
    "plt.ylabel(\"feature\")\n",
    "plt.figure(figsize=(5,5))\n",
    "fig.set_size_inches(10, 8, forward=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "1feb4777",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_clean = X.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "9cebbc3f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1709857, 45)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_clean.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "e052303f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(45,)"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "01921e72",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_mock = X_clean.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "8c7c8f17",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1709857, 44)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_mock = np.delete(X_clean, 5, 1)\n",
    "X_mock.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "ce8eb532",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.08455971370712748, 0.25569036308079385, 0.0008847651317850527)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features[0], features[1], features[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "920ef6a5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.029621723015000694"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features[20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "b0193aee",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "45"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "4dc573c5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[2,\n",
       " 3,\n",
       " 4,\n",
       " 5,\n",
       " 6,\n",
       " 7,\n",
       " 8,\n",
       " 9,\n",
       " 11,\n",
       " 12,\n",
       " 13,\n",
       " 14,\n",
       " 15,\n",
       " 16,\n",
       " 17,\n",
       " 18,\n",
       " 19,\n",
       " 21,\n",
       " 22,\n",
       " 23,\n",
       " 24,\n",
       " 25,\n",
       " 26,\n",
       " 27,\n",
       " 28,\n",
       " 29,\n",
       " 30,\n",
       " 31,\n",
       " 32,\n",
       " 33,\n",
       " 34,\n",
       " 35,\n",
       " 41,\n",
       " 42,\n",
       " 43,\n",
       " 44]"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "deletable = []\n",
    "for i in range(0, len(features)):\n",
    "    if features[i] < 0.025:\n",
    "        deletable.append(i)\n",
    "#         X_clean = np.delete(X_clean, i, 1)\n",
    "deletable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "2ba95c16",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_clean = np.delete(X_clean, deletable, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "aa4520d9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1709857, 9)"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_clean.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "a8d65375",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_clean_train, X_clean_test, y_clean_train, y_clean_test = train_test_split(X_clean, y, test_size=0.3, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "a6ffc13d",
   "metadata": {},
   "outputs": [],
   "source": [
    "skf_clean = StratifiedKFold(n_splits=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "3442d513",
   "metadata": {},
   "outputs": [],
   "source": [
    "rfc_clean = RandomForestClassifier()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "5ad04359",
   "metadata": {},
   "outputs": [],
   "source": [
    "random_search_clean = RandomizedSearchCV(rfc_clean, param_grid, cv=skf_clean, verbose=2, random_state=42, n_jobs = -1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "16603c66",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 2 folds for each of 1 candidates, totalling 2 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.9/site-packages/sklearn/model_selection/_search.py:292: UserWarning: The total space of parameters 1 is smaller than n_iter=10. Running 1 iterations. For exhaustive searches, use GridSearchCV.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "ename": "AttributeError",
     "evalue": "'RandomizedSearchCV' object has no attribute 'cv_results_'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "Input \u001b[0;32mIn [63]\u001b[0m, in \u001b[0;36m<cell line: 2>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m random_search_clean\u001b[38;5;241m.\u001b[39mfit(X_clean_train, y_clean_train)\n\u001b[0;32m----> 2\u001b[0m \u001b[38;5;28mprint\u001b[39m(\u001b[43mrandom_search\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcv_results_\u001b[49m)\n",
      "\u001b[0;31mAttributeError\u001b[0m: 'RandomizedSearchCV' object has no attribute 'cv_results_'"
     ]
    }
   ],
   "source": [
    "random_search_clean.fit(X_clean_train, y_clean_train)\n",
    "print(random_search.cv_results_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "20810c7f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9709958817693625"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "random_search_clean.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "989f52e2",
   "metadata": {},
   "outputs": [],
   "source": [
    "classifier_clean = random_search_clean.best_estimator_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "1843a89e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9990308288335106"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classifier_clean.score(X_clean_train, y_clean_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "cc6e62da",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.982327987866453"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classifier_clean.score(X_clean_test, y_clean_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "a62a500c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 360x360 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#modelname.feature_importance_\n",
    "features_clean = classifier_clean.feature_importances_\n",
    "ylabels = ['Weekday', 'Seconds', 'M008', 'M018', 'T001', 'T002', 'T003', 'T004', 'T005']\n",
    "#plot\n",
    "fig, ax = plt.subplots()\n",
    "width = 0.4 # the width of the bars\n",
    "ind = np.arange(len(features_clean)) # the x locations for the groups\n",
    "ax.barh(ind, features_clean, width, color=\"green\")\n",
    "ax.set_yticks(ind+width/10)\n",
    "ax.set_yticklabels(ylabels, minor=False, fontsize=16)\n",
    "plt.title(\"Feature importance in RandomForest Classifier\", fontsize=16)\n",
    "plt.xlabel(\"Relative importance\", fontsize=16)\n",
    "plt.ylabel(\"Feature\", fontsize=16)\n",
    "plt.figure(figsize=(5,5))\n",
    "fig.set_size_inches(8, 6, forward=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2a9d4565",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}